They did not report their methods for the ecig sample acquisition. It's quite possible that they were not using the vaporizer under the normal conditions. There was a previous study where they were running the device at excessive voltages over tens of seconds, causing pyrolysis which would never happen under normal usage. Simply because it would taste awful and make the user cough a lot.

I would have liked to see control experiments: analyse the e-liquid without vaporization, and analyse pure glycerol/propylene glycol. The former may show a response, but the latter should not show any at all.